Research on the cloud platform resource management technology for surveillance video analysis
Yonglong Zhuang1,Xiaolan Weng2,Xianghe Wei2
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12B) 60-65
1Modern Educational Technology Center, Huaiyin Normal University, Jiangsu Huai'an 223300, China
2College of Computer Science and Technology, Huaiyin Normal University, Jiangsu Huai'an 223300, China
As the cloud computing provides the characteristics of the computing resources which can be randomly used or rented, it has become a topic issue to develop a method for dynamically adjusting the computing resources in terms of the service. The paper introduces a method for dynamically adjusting the computing resources on the basis of the surveillance video analysis result. The contents of the surveillance video in the method are cut into many segments, distributed into many computing nodes and to be analyzed. In the meantime, the estimated method for the workload can be sensed through the introduced contents and the system workload can be predicted. The numbers of the computing nodes can be dynamically adjusted without affecting the quality of the service, and the utilized computing resources can be reduced to the minimum as soon as possible. The experimental result shows that the method introduced by the paper can effectively predict the system workload, and can show its advantages in the computing cost and work completion under the situation of not affecting the service quality.